Tax Loss Harvesting and Standby Substitutes
The practical challenge when tax loss harvesting is maintaining a continuous asset class exposure at target levels without time gaps, while avoiding penalties under the IRS Wash Sale Rule (IRC Section 1091).
The problem with time gaps is that significant market moves can occur in the 30-day waiting period of the Wash Sale rule, which would prevent the portfolio from achieving the risk and return expectations on which the portfolio asset allocation was designed.
The solution to the problem is substitution. Immediately upon realizing a loss in one fund, open a position in an alternate fund that is similar to, but not "substantially identical" to, the fund on which the loss was realized.
After the waiting period of 30 days, close the substitute fund position and reopen the original position (assuming the alternate fund is a second best choice). Or, if the substitute fund is equally attractive for the portfolio purpose, keep the substitute position and make the original your new alternate substitute for use with a potential future tax loss harvest.
Importance of a Standby Substitution List:
You should develop a tax harvesting fund substitution list now, not later. To do it right, you need some time to identify prospects, research their particulars (including whether they have enough liquidity), and make a determination as to their suitability for your purposes and for Wash Sale Rule compliance. You won't have time for that if you try to do it on the fly when you decide to realize a loss. Make the list when you have time to do it well.
Here is a sample of a list for some key asset classes.
The problem with advice. You should consult your tax advisor. Everybody says that. Your stock brokerage or mutual fund company says that. The IRS help line says that. And, we say that -- although we are not afraid to tell you what we think, so long as you don't act on it without running it by your tax advisor.
There is a high probability that your tax advisor will bounce you back to your stockbroker or mutual fund company. That's because the regulations are vague. Unless you manage a wide path around the Wash Sale Rule, your tax advisor may not know enough about the particular investments you are using to say with comfort whether or not they comply.
The solution to the tax advisor problem is preparation. We are investment managers, not tax advisors, but we know far more about the investment funds in our portfolios than our tax advisors. If you are managing your money, you should know more about your investments too.
If you understand the Wash Sale Rule principals and terms, and you have done thorough research on your investment funds, you will be in a good position to present your logic to your tax advisor, who will then be able to apply tax expertise to your investment fund knowledge to make appropriate decisions.
We recommend establishing guidelines with your tax advisor to be used before you invest, rather than debating the issues on a transaction basis after you invest.
What's in this article?:
This article explores solutions for fund-based portfolios, and provides you with some of the preparation and ammunition to have the necessary informed discussion with your tax advisor.
Better yet, we recommend presenting a list of funds and their "not substantially identical" substitutes to your tax advisor to achieve a higher level of comfort with your compliance. We present some examples in this article that we feel are compliant.
Summary of the Wash Sale Rule:
A tax loss deduction is disallowed if, in the 30-day period before or after a realized loss, the investor replaces the security on which the loss was realized with another security that "appears" to be "substantially identical".
IRS Publication 550 discusses the Wash Sale Rule and provides some examples, but does not discuss or provide examples for investment funds.
"Appears" from the statute is not defined (and is not mentioned in Publication 550). Obviously the burden of proof is on the investor, which means the investor should be well armed with fund characteristics and attributes to argue the matter of "appearances" if a deduction is challenged by the IRS.
Publication 550 says this about "substantially identical":
"In determining whether stock or securities are substantially identical, you must consider all the facts and circumstances in your particular case. Ordinarily, stocks or securities of one corporation are not considered substantially identical to stocks or securities of another corporation. However, they may be substantially identical in some cases."
As you have just read, "substantially identical" is not defined -- investor beware. However, based on the opinions of a variety of tax experts, who in turn have relied on case law, we think a general consensus could probably found with something like:
Substantially identical means the same in all important respects, particularly economic position and risk exposure, but not correlation of return, except for the case of two companies undergoing merger or the case of one security convertible to another.
In the end, we think the arguments will revolve around differences in your "economic position" and "risk exposure" between investing in the original fund versus investing in the substitute fund. Be prepared to argue those differences.
Conservative Substitution Working Rules:
In an earlier article examining tax loss harvesting as it relates to the IRS Wash Sale Rule, we concluded with this set of working rules for replacing a fund during the 30-day exclusionary period before and after the date of the realized loss to satisfy the question as to whether there is the "appearance" that the funds are "substantially identical":
1) Replace one actively managed fund for another if the issuers and managers are not the same.
2) Replace one index fund for another index fund, even though there may be overlap in their holdings, if the index of the sold and of the purchased fund are not the same index, and the differences between the indices are obviously significant (e.g. either they have a significantly different number of index constituents or they hold the same constituents but in significantly different proportions).
3) Replace any index fund with any actively managed fund regardless of issuer
4) Replace any actively managed fund with any index fund regardless of issuer.
5) Ask your "personal tax adviser" if they agree before adopting these working rules -- no statement in this article is personal tax advice.
In the graphic above, the type "A" Substitutes for the Example Original funds all comply with these working rules. We think of them as conservative substitutes, because we believe they neither "appear to be" nor are they "substantially identical".
Those substitutes in comparison broadly represent the asset class of the originals they replace, but they:
are sponsored by different investment management firms
track different indices
reflect that the indices have different numbers of index constituents or formula proportions of constituents
have different management expense ratios
yield different cash distributions (except for the commodity funds with no distributions)
hold different numbers portfolio holdings
hold different proportions for all or most of the holdings they have in common.
For these reasons, we think the examples given are highly defensible, because of the number of differences and the obviousness of their significance to the "economic position" and "risk exposure" associated with them.
Aggressive Substitution Working Rule:
We think it should be possible to use a substitute fund that has the same index benchmark target as the original fund, if the "facts and circumstances" show the funds to be nominally the same, but factually different.
The first obstacle to this aggressive approach is the superficial "appearance" of being identical. The second obstacle is how much difference the IRS will require to accept that the two index funds are not substantially identical.
This area is untested and the issues are not obvious to someone who does not delve into investment details. That makes it aggressive (risky). It's got to be tested sometime, so we'll start the debate now with this article.
Here is the aggressive rule, after which we will explore factual issues for some pairs to see the spectrum of difference and no-difference that can exist:
Replace one index fund for another index fund offered by different sponsoring companies, but each based on the same index, if the way the two funds pursue the same index is not substantially identical in terms of attributes such as:
index representative sampling versus replication
major differences in the number of holdings used for representative sampling
management expense ratios (and sales or redemption loads, if any)
cash distribution yield of income and of capital gains
portfolio turnover rate
use of derivatives versus no use of derivatives to track index
credit quality, average duration or maturity for bond funds
valuation ratios (such as P/E, P/B, P/S, and P/CF) for equity funds
position weights of largest portfolio holdings
The Case of Four Pairs:
Let's look at a spectrum of four pairs of funds each tracking the same index to see how differences can be found:
- SPY : IVV
- EFA : VEA
- VWO : EEM
- BND : AGG
SPY : IVV
This pair is substantially identical. They would not work for substitution under the Wash Sale Rule. They both track the S&P 500 index. They both use replication (each holds all 500 constituents in exact index weights). They have negligible differences in expense ratios. All of the portfolio characteristics are therefore essentially identical as are distributions. [number of different attributes = 0]
EFA : VEA
This pair may not be substantially identical, although they are close the edge. Their greatest differences are in expense ratio (reflected directly in yield), the number of holdings and the weight of "Other/Not Classified" holdings. [number of different attributes = 4]
VWO : EEM
This pair is not substantially identical when you look into the portfolio attributes. They have different average market-caps, different P/CF, massively different expense ratios, different yields beyond the effect of expense ratios, different weights in the top 10 holdings, different regional weights, and different numbers of holdings. [number of different attributes =7]
BND : AGG
This pair is not substantially identical in terms of portfolio attributes. While they both use representative sampling, the difference between 172 portf0lio positions and 11, 930 is enormous (and the risk of tracking error is greater in one than the other). The average effective maturity is different and the average credit quality is different (both classical indicators of relative risk). The weight of positions in the top 10 is widely different. The SEC yield is different. The expense ratio of one is nearly double the other (expenses being extra critical to outcomes in a fixed income fund). The distribution of bond holdings by credit quality is different. [number of different attributes = 7]
If you stick to the conservative working rules, you should have a fairly short meeting with your tax advisor. If you want to use the aggressive rule, arm yourself with plenty of data and be prepared to receive a larger invoice for professional services rendered.
If you have other views on the matter, please do comment as we all learn about this uncharted area together.